This investigation proposes to identify protective factors to prevent and/or buffer youth perpetration of aggression, delinquency, and violent behaviors. Data will be drawn from the Longitudinal Studies of Child Abuse and Neglect (LONGSCAN). LONGSCAN is the largest prospective study of 1,354 children who have been maltreated or who were thought to be at risk for maltreatment. Children in this cohort have been carefully followed since age 4 with extensive data collected every two years until age 18. Aggressive behavior and delinquency have been carefully measured repeatedly from multiple informants. This dataset is ideal for examining the pathways to delinquency that may be linked to a history of child maltreatment and witnessed violence and for examining protective factors that offset those pathways. Data from ages 4 through 16 will serve as the study sample for these proposed analyses. Data collection has been completed through age 14, and will be completed for the age 16 interviews by the beginning of 2009. Currently, 1,135 participants have completed one or more of the targeted interviews (12, 14, and 16), comprising 84% of the baseline sample. The gender distribution for the baseline sample, and for the corresponding analysis sample, is approximately equally distributed. Slightly over half of the baseline sample is African American (55.2%), followed by Caucasian (25.0%), mixed race (11.2%), Latino/a (7.2%), and other (1.4%). Study aims for this proposal are (1) determine the stability of risk and protective factors and the ability of factors to offset risks in reducing the perpetration of youth violence at three time points measured during the adolescent period, (2) examine the cumulative buffering effect given the addition of each protective factor, and (3) for those youth already engaging in violence and aggression, determine which factors increase the likelihood of desistence at ages 14 and 16. Proposed protective factors include supportive adult, mandated substitute care, social skills, academic commitment, neighborhood social bonding and informal social control, parental monitoring, and parent-child relationship quality. Risk factors include child maltreatment, exposure to violence, early aggression and delinquency. The utilization of the LONGSCAN data to examine risk and protective factors of maltreated youth affords and unique and powerful opportunity to explore these factors within and across levels of the socio-ecological model at multiple time points during adolescence. PUBLIC HEALTH RELEVANCE: Violence is a serious public health problem facing every sector of American society, disproportionately affecting children, adolescents and young adults. We propose to analyze data already collected from the Longitudinal Studies of Child Abuse and Neglect (LONGSCAN) to identify factors that compensate and/or modify the effects of risk on adolescent perpetration of violence and aggression in a high risk sample of youth. This dataset is ideal for examining the pathways to delinquency that may be linked to a history of child maltreatment and witnessed violence and for examining protective factors that offset those pathways.